Most units finish an audit the same way: a spreadsheet, a compliance percentage, and a slide that says "we will re-educate staff." A few weeks later the number has not moved. The gap is rarely the data. It is the absence of a structured plan that turns the finding into a small, testable change at the bedside.
Audit data tells you what is happening. An improvement plan answers what we will do differently, on which patients, starting Monday. This article walks through moving from a chart audit or quality report to a practical, evidence-based plan you can run on your own unit, using the Model for Improvement and Plan-Do-Study-Act (PDSA) framework that AHRQ and the Institute for Healthcare Improvement recommend.
Start with the three questions, not the intervention
Before designing any fix, AHRQ's quality improvement guidance frames the work around three questions from the Model for Improvement: what are we trying to accomplish, how will we know a change is an improvement, and what changes can we make that will result in improvement. Answer them in that order.
Audit data feeds the first two questions. Suppose your fall audit shows that hourly rounding was documented on 58 percent of charts. The aim is not "improve rounding." A useful aim is specific, measurable, and time-bound: increase documented purposeful rounding on the night shift of unit 4 East from 58 to 90 percent within eight weeks. Notice the narrowing. One shift, one unit, one behavior. That scope is deliberate and it is what makes a plan testable.
For the second question, decide your measure before you act. Reuse the same audit definition you already have so the before and after are comparable. If the baseline counted any rounding note in the past two hours, keep that exact definition. Changing the measure mid-stream is one of the most common ways teams fool themselves into seeing improvement that is not real.
Turn the finding into a change idea you can test small
The third question, what change will produce improvement, is where audit data points you toward a theory, not just a target. Ask the bedside why. If rounding is not documented, is the rounding not happening, or is it happening but not charted? Those are two completely different problems with different fixes. A quick look at a handful of charts, plus asking the nurses on that shift, usually tells you which.
From that theory, pick one change idea. Examples that stay within nursing scope and defer to facility policy:
- A rounding prompt built into the existing handoff, not a new form.
- Repositioning the rounding log to the point of care so charting happens in the room.
- A peer reminder during the 0200 lull, when the audit showed the biggest gap.
Resist bundling five changes at once. AHRQ describes the second half of the Model for Improvement as rapid cycle improvement: testing interventions on a small scale, learning, then modifying for the next cycle. If you change five things and the number moves, you will not know which one worked, and you will not be able to spread it.
Run the PDSA cycle and document each one
The IHI PDSA Worksheet structures a single test of change into four steps: Plan the test, Do the test by carrying it out, Study by observing and learning from the results, and Act by deciding what to modify. Two practices from that tool matter most for nurses.
First, predict the result before you run the test. Write down what you expect rounding documentation to look like after one week. A prediction that misses tells you your theory was wrong, which is far more useful than a vague sense that things feel better.
Second, fill out one worksheet per test and keep them all. The IHI guidance is explicit that each change goes through several PDSA cycles and that teams should keep a file of every worksheet. That file becomes your evidence trail when you bring results to the council or the manager, and it is what lets the next unit replicate what you did.
A change idea is a hypothesis. A PDSA cycle is the experiment that tells you whether the hypothesis held on real patients.
Why discipline beats enthusiasm
It is tempting to skip the small test and roll a promising idea out unit-wide. The evidence cautions against it. A systematic review of 120 PDSA quality improvement projects found that while 98 percent reported improvement, only 27 percent met their pre-specified aim, and just 4 percent adhered to all four core methodological features. The features most often skipped were exactly the ones that protect you from a false positive: small-scale testing, iterative cycles, continuous data collection, and a stated rationale for why the change should work.
The practical takeaway for the bedside is reassuring. You do not need a statistician or a large project. You need a narrow aim, a measure you do not move, one change tested small, a written prediction, and an honest look at the next batch of charts. When a cycle works, expand the population or the shift and test again. When it does not, you have learned something cheaply and changed nothing for patients who were counting on it.
Bring the plan, not just the percentage, to your shared governance council or quality team, and align it with your facility's existing QI structure before you begin.